ASUS GeForce GTX 590 3 GB

[MxPhenom 216]

i run a lot more voltage through other cards during testing and this never happens .. on gtx 590 with nvidia's power capping feature which is designed for that purpose it doesnt work. i think it's my obligation to tell you, no ?

could that be fixed through drivers?? i feel like the 590 could be an amazing card if the VRM we actually good and if they didnt run the 2 15--200w GPUs on 4+1 phase power. thats obviously not enough. they could have added half an inch or so of PCB and added some phases. Price the card a tiny bit higher, but then it would be king.

 

[Airbrushkid]

Why have others not had the same thing happen during there testing?

 

[TheMailMan78]

could that be fixed through drivers?? i feel like the 590 could be an amazing card if the VRM we actually good and if they didnt run the 2 15--200w GPUs on 4+1 phase power. thats obviously not enough. they could have added half an inch or so of PCB and added some phases. Price the card a tiny bit higher, but then it would be king.

Its not a driver issue. Its a design flaw.

 

[qubit]

could that be fixed through drivers?? i feel like the 590 could be an amazing card if the VRM we actually good and if they didnt run the 2 15--200w GPUs on 4+1 phase power. thats obviously not enough. they could have added half an inch or so of PCB and added some phases. Price the card a tiny bit higher, but then it would be king.

Yeah, my Gigabyte GTX 560 Super Overclock I recently bought (and returned) had 7 or 8 phases onboard and that card is very much weaker than the 590. It looks like nvidia cheaped out there. btw, I dunno how many phases my new Zotac GTX 580 has, I'd have to look it up. I do know that it doesn't squeal too loudly though, if I let it freewheel at 500fps+ (but not for long, to avoid damage).

Its not a driver issue. Its a design flaw.

Yup, exactly.

 

[cowie]

the card is advertised as that being a supported feature. Also, nvidia claimed to have drivers with power throttling to prevent that exact situation.

cards are not supported for 1.2 anything just 1.05v in ab.
some have said this setting is 1.10v with dmm
still its better to have guys that dont pay for them blow them up to give a heads up to anyone that does buy these cards.
sofar not one paying customer has reported killing there card.......yet

 

[W1zzard]

cards are not supported for 1.2 anything just 1.05v in ab.
some have said this setting is 1.10v with dmm
still its better to have guys that dont pay for them blow them up to give a heads up to anyone that does buy these cards.
sofar not one paying customer has reported killing there card.......yet

nobody has paid for a gtx 590 yet .. what would you say if those cards started dying left and right and i have said in my review "oh it will work great with 1.2v, it's perfectly safe" ?

 

[cowie]

nobody has paid for a gtx 590 yet .. what would you say if those cards started dying left and right and i have said in my review "oh it will work great with 1.2v, it's perfectly safe" ?

wait do they make you pay for those? i hope not.
you doing your job to find these things out for us thats why i wait to read your reviews.
you would never say that becuase that would be a foolish thing to garrenty. we all (should)know that overclock and overvolting is risky.
ab is limited for the end users to 1.05v, how did you get more or why did you try more? if you did i mean.

 

[Mussels]

Why have others not had the same thing happen during there testing?

it has. the youtube video has been linked a few times by now.

 

[Airbrushkid]

What do you mean linked? If they cannot make video of there own findings and have to use others. Then I cannot believe it.

 

[wahdangun]

great review wizz and shame on nvdia, WTF they cheapen out on this premium extreme high end card ???

btw wizz do you thing you will get another card for quad SLi review ??

 

[W1zzard]

btw wizz do you thing you will get another card for quad SLi review ??

nvidia is sending me a second card on monday, just got the info from them

 

[qubit]

nvidia is sending me one on monday, just got the info from them

And after benching SLI performance, will you try to overvolt it again and see if it's any better, or is that off limits now?

 

[W1zzard]

And after benching SLI performance, will you try to overvolt it again and see if it's any better, or is that off limits now?

i could offer nvidia to do some new testing with their new driver, i'll look into that after the quad sli review (not gonna break another card before the quad sli review is posted)

 

[wahdangun]

nvidia is sending me a second card on monday, just got the info from them

i could offer nvidia to do some new testing with their new driver, i'll look into that after the quad sli review (not gonna break another card before the quad sli review is posted)

wow, thats great . i can't wait for your quad SLi review, and please do more testing like upping the frequency to the limit and check it stability and also check if the new driver really solve the pooping problem lol

 

[HXL492]

just wondering if this is the easter egg

 

[laszlo]

Its not a driver issue. Its a design flaw.

srry but i disagree

in my opinion due gpu design (2 high leaking which result in 2 much power draw ) with stock frequency & the other components used specifically for this card in order to minimize cost, had the result of obtaining this card who had a fragile balance between components/clocks/power draw

if you break the balance 2 much some component fail and bye bye warranty

as i saw also nvidia don't encourage the o/c of this as they knew it from beginning but they don't care as no no warranty for this blow-ups is accepted;i'm sure all cards work flawlessly without o/c;basically buy it but is not our problem if you oc is your wallet

 

[ERazer]

What do you mean linked? If they cannot make video of there own findings and have to use others. Then I cannot believe it.

tell me do you video every bench do you do? nobody knew some of this card gonna blow, if wiz knew his gonna blow hell prolly made an awesome 3d hd movie out of it, id paid to see that

 

[newtekie1]

Its not a driver issue. Its a design flaw.

I disagree, it is a driver issue preventing the power limitter from working properly.

I also think it is a driver issue that prevents the power limitter from lowering the voltages as well as the clocks. Which means that, when a user overvolts, even at the "safety" clocks the card is pulling too much power.

I think nVidia should just limit the maximum voltage to 1.00v or 1.10v via the BIOS, like they did with the GTX400 cards which were limitted to 1.087v. That way the likelyhood of someone popping a card is a lot lower because it seems like 1.2v is the point where these tend to pop. If someone wants to increase the voltage beyond 1.10v they can edit the BIOS, and then it is on them for raising the maximum, not nVidia's fault.

 

[wahdangun]

I disagree, it is a driver issue preventing the power limitter from working properly.

I also think it is a driver issue that prevents the power limitter from lowering the voltages as well as the clocks. Which means that, when a user overvolts, even at the "safety" clocks the card is pulling too much power.

I think nVidia should just limit the maximum voltage to 1.00v or 1.10v via the BIOS, like they did with the GTX400 cards which were limitted to 1.087v. That way the likelyhood of someone popping a card is a lot lower because it seems like 1.2v is the point where these tend to pop. If someone wants to increase the voltage beyond 1.10v they can edit the BIOS, and then it is on them for raising the maximum, not nVidia's fault.

the problem is some vendor (like asus and evga) encourage overvolting (hell even in boxes they said it can be increased up to %50) so its design flaw (because the weak VRM)

 

[cowie]

I agree newtekie 0.95v -1.21v is over 25% increase in voltage,thats alot.

 

[Airbrushkid]

tell me do you video every bench do you do? nobody knew some of this card gonna blow, if wiz knew his gonna blow hell prolly made an awesome 3d hd movie out of it, id paid to see that

If I was running a review site and the companies where sending my product review. Then I would be making videos of the reviews!

 

[W1zzard]

If I was running a review site and the companies where sending my product review. Then I would be making videos of the reviews!

you can do that when you have a review site, there are many reasons for and against video.

for tpu we chose to go the non video route, and apparently it's working well, going by the amounts of traffic we get

 

[cadaveca]

Yeah, you guys don't want to see my ugly mug and mitts for mobo reviews...trust me on that one.

Doing video for reviews exponentially increases the amount of work required to do a review. Gotta do the whole review, plus video/audio editting...more hosting and bandwidth requirements, more possible problems, too.

 

[newtekie1]

the problem is some vendor (like asus and evga) encourage overvolting (hell even in boxes they said it can be increased up to %50) so its design flaw (because the weak VRM)

They say "up to 50% Faster" they don't say you can increase the voltage 50%. They can limit the voltage to 1.1v and still encourage overvolting.

However, that isn't the issue, the issue is that the powerlimitter in the driver is broken.

 

[the54thvoid]

whats the max overclock then like 675??

Head over to Hexus next week and find out

http://www.hexus.net/content/item.php?item=29753

POV TGT Ultra charged is clocked at 691Mhz Core. Substantially up from 607Mhz.